The goal of this research program is to construct a model of human red/green color-vision defects that is based upon the molecular biology of the color-vision genes. In the April 11, 1986 edition of Science, Nathans and colleagues (including one of the authors of this proposal (Piantanida)), reported (1) the nucleotide sequences of the three genes that encode the human cone photopigments. (2) the existence of multiple genotypes among color-normal males, and (3) the occurrence of hybrid genes (and often one or both normal color- vision genes) in persons with red/green color-vision defects (Nathans et al, 1986 a,b). The proposed research project would continue that investigation of the molecular genetics of red/green color-vision defects, concentrating initially on sequencing the hybrid genes, and eventually producing the first comprehensive DNA-based model of normal and defective human color vision. The hybrid genes, which appear to have arisen by unequal intragenic recombination of homologous regions of the red-and green-cone genes, were identified in the studies cited above by polymorphism of their restriction fragment lengths. However, that method does not provide sufficient resolution to determine the exact point of recombination. Because the hybrid genes appear to be the key to understanding red/green color-vision defects, the proposed research will concentrate on specifying their nucleotide sequences. DNA samples from color-defective males will be digested with restriction enzymes and the restriction fragments analyzed by Southern blotting. A determination of genotype will be made by stoichiometric analysis of the Southern blots, and restriction fragments of interest will be cloned into lambda vectors, identified by plaque hybridization, subcloned into an M13 vector, and sequenced by the dideoxy method. In addition by hybrid genes, the genotypes of some persons with red/green color-vision defects include a variable number of ostensibly normal color-vision genes. More generally, the results of Nathans et al. (1986a) show that color-vision genotypes are not related to the six traditionally-defined defective red/green color- vision phenotypes in a simple manner. The proposed research is aimed at clarifying the relationship between color-vision genotypes and phenotypes. Specifically, it will (1) explore further the number and types of genes found in color-normal and color- defective persons in an attempt to develop a complete genotype catalog; (2) expand the red/green color-vision phenotype catalog based on existing psychophysical data so that subjects can be sorted into more precise phenotypes by pseudoisochromatic plate tests, small- and large-field anomaloscopy, neutral point determination, and other psychophysical tests; and (3) most important, develop a model of defective human red/green color vision based upon a comparison of genotypic and phenotypic features, which will have important implications for models of normal color vision.